Basis set compatibility issue

Questions
#1

I am attempting to reproduce the SOCC calculations with SF-TDDFT/TDA as described in the paper (DOI: 10.1063/5.0130868), but have encountered two basis set compatibility issues:

  1. For the results in Table VI, the cc-pcVTZ basis set is not supported for the Br atom. This is somewhat expected, as the basis set is not defined for Br in either the Basis Set Exchange (BSE) or the Q-Chem manual.

  2. For the results in Tables VII and VIII, the cc-pVDZ basis set is unexpectedly reported as “Basis not supported for the above atom” when used with the Fe atom. This is confusing because cc-pVDZ is explicitly listed as supported for Fe in both the Q-Chem manual and BSE.

I would greatly appreciate guidance on whether it is possible to reproduce the Fe-atom calculations with the cc-pVDZ basis set, or if there is a workaround for this compatibility issue.

Thank you!

Input for the Fe case:
$molecule
0 1
Fe 0.0000091774 -0.0225847387 -0.0024111887
$end

$rem
method = PBE0
basis = cc-pvdz
cis_n_roots = 10
spin_flip = true
calc_soc = 2
wang_ziegler_kernel = true
$end

Output ():
You are running Q-Chem version: 6.2.1-1
…
Q-Chem 6.2, Q-Chem, Inc., Pleasanton, CA (2024)
…
Q-Chem 6.2.1 for Intel X86 EM64T Linux
…
Requested basis set is cc-pVDZ

Unexpected End of File for file

/home/peter/qchem/qcaux/basis/cc-pVDZ.bas

was found while looking for Atomic Symbol Fe.

Check to see if this file has a basis for that atomic symbol.
The input format searched for here should be:
<0>
Q-Chem fatal error occurred in module forms1/AtomicBasis.C, line 1157:

Basis not supported for the above atom.

#2

For the first question, cc-pVTZ is definitely defined for Br, in both the Basis Set Exchange and in the latest version of Q-Chem (v. 6.4). Some of the Dunning basis sets were updated a few years ago so possibly very old versions of Q-Chem might not have all of the atoms that are currently in the BSE, in which case you can use BASIS = GEN and a $basis section.

For the second question, use cc-pVDZ-full, which I believe extends cc-pVDZ to the 3d metals.

#3

Thank you very much for your prompt and helpful response!

Regarding the first issue, I would like to clarify that the basis set I initially encountered compatibility problems with was cc-pcVTZ (not cc-pVTZ) for the Br atom. But cc-pVTZ should also work for my reproduction efforts, and I will consider it as a feasible alternative.

For the second issue, your recommendation to use cc-pVDZ-full has fully resolved the problem with the Fe atom calculations.

#4

Recent versions of Q-Chem do have cc-pCVTZ but the Basis Set Exchange suggests that this basis set is only supported for elements through Ar (and also Ca), which is consistent with what’s in Q-Chem’s library.